Emergent rate-based dynamics in duplicate-free populations of spiking neurons

2023/03/09 by Valentin Schmutz, Johanni Brea, Schmutz, Valentin +3 · 1 voice
Biochemistry, Genetics and Molecular Biology · #FOS: Biological sciences #Neurons and Cognition (q-bio.NC) #q-bio.NC

paper · pdf · doi:10.48550/arxiv.2303.05174

arxiv published 2023/03/09 · arxiv updated 2024/11/07

Abstract

Can Spiking Neural Networks (SNNs) approximate the dynamics of Recurrent Neural Networks (RNNs)? Arguments in classical mean-field theory based on laws of large numbers provide a positive answer when each neuron in the network has many "duplicates", i.e. other neurons with almost perfectly correlated inputs. Using a disordered network model that guarantees the absence of duplicates, we show that duplicate-free SNNs can converge to RNNs, thanks to the concentration of measure phenomenon. This result reveals a general mechanism underlying the emergence of rate-based dynamics in large SNNs.

Discussions

• Dear Rubén, Would love to be added as well. Here is a recent neurotheory preprint of mine arxiv.org/abs/2303.05174 (to appear in PRL). Thanks a lot! [bsky, 4 points, 1 comments]

Related

• Replica-mean-field limits for intensity-based neural networks
• Minimizing information loss reduces spiking neuronal networks to differential equations
• Modeling Higher-Order Interactions in Sparse and Heavy-Tailed Neural Population Activity
• Stochastic dynamics of a finite-size spiking neural network
• Spontaneous Dynamics of Asymmetric Random Recurrent Spiking Neural Networks